How Should a SaaS Company Ask How Did You Hear About Us?

How should a SaaS company ask "how did you hear about us"?

Ask it as an open text field on your highest intent form, usually the demo request or signup, and make it required only if your volume is low. Then code the answers into a short list of channels every week. Open text captures the podcast, the Slack group, or the AI answer that a dropdown would never list.

Most B2B SaaS teams already have this question somewhere. The problem is how it is asked. A dropdown with ten generic options, placed on every form, filled in by people who just want to get to the next screen, produces data nobody trusts. Then the team stops looking at it.

I build websites and the automations behind them for lead generation, and this one field comes up in almost every project. Here is how I set it up so the answers are actually useful to a founder or a marketing lead deciding where to spend next quarter.

Why ask buyers directly when you already have analytics?

Because analytics only sees clicks, and buyers rarely decide on a click. A prospect might hear about you on a podcast, see a LinkedIn post, ask ChatGPT about your category, and then search your name. Analytics records the last step as direct or branded search. Only the buyer can tell you what started it.

This gap has grown as more research happens in places that do not pass a tracking link. Private communities, word of mouth, voice notes, and AI assistants all shape buying decisions without leaving a trail in GA4 or HubSpot.

Software attribution is still useful. It tells you which pages people saw and which campaigns brought clicks. Self-reported attribution answers a different question: what made this person aware of you and willing to look. You want both, and you want to know which question each one answers.

Whether you can trust what buyers say is its own debate. I covered that in should you trust self-reported attribution. This article assumes you have decided to ask, and focuses on asking well.

Should the field be a dropdown or open text?

Open text, in almost every case. Dropdowns force buyers into your categories, and they tend to pick the first plausible option. Open text lets them name the actual source, like a specific newsletter, a person, or an AI tool. The extra work of reading answers is the price of data that reflects reality.

The dropdown problem is subtle. You build the list from channels you already know about, so the list can never surprise you. If a niche community is quietly sending your best buyers, a dropdown will hide it under "Other" or "Social media".

Open text answers are messy, and that is fine. "My friend at a fintech startup" and "saw your founder on a podcast" tell you more than a tidy category ever could. You will clean them up afterwards, which I explain below.

If your form volume is very high, a hybrid can work: a short dropdown plus an optional "tell us more" text field. But for most early and mid stage SaaS companies, form volume is low enough that reading every answer is realistic and valuable.

Where should the question appear?

Put it on the form that signals real buying intent, such as a demo request, a sales contact form, or the final step of a trial signup. Skip it on newsletter signups and content downloads. Asking on low intent forms adds friction where you need the least, and the answers matter less there.

Placement within the form matters too. I put it near the end, after the fields a buyer expects to fill in. Someone who has typed their name, work email, and company is already committed, and one more short question rarely stops them.

For product-led SaaS with self-serve signup, the onboarding flow is often a better home than the signup form. A single question on the first screen after account creation catches people at a moment when they are curious, not rushing.

If you are reviewing your demo form at the same time, my piece on what a demo request form should ask covers the other fields and which ones to cut.

Should the field be required?

Make it required when form volume is low and every answer matters, and optional when volume is high and friction costs you leads. A required field produces more answers but also more junk like "Google" or a single letter. An optional field produces fewer answers, but those answers tend to be more thoughtful.

My default for most B2B SaaS demo forms is required, with a clear label and a short hint underneath. Something like "How did you first hear about us? A person, podcast, community, or tool is perfect." The hint teaches buyers what kind of answer you want.

Watch the results for a few weeks. If you see many throwaway answers, the wording needs work before the requirement does. If you see form completion drop, test making it optional and compare the quality of what you get.

How do you turn messy answers into something you can report on?

Code them. Once a week, read every new answer and assign it to one channel from a short, fixed list, such as podcast, community, referral, LinkedIn, AI assistant, search, and event. Store the code in a separate field from the raw answer. Then report on the codes and keep the raw text for detail.

Keep the raw answer forever. The coded channel is for charts, but the raw text is where insight lives. "Referral" in a chart tells you little. Reading ten referral answers might show that three came from the same customer, which tells you who your best advocate is.

Keep the code list short and stable. If you add new categories every week, your trend lines break. I start with seven or eight codes and only add one when a new source keeps appearing.

The coding can be manual, partly automated, or done with an AI model suggesting a code that a person confirms. Whatever you use, a human should check the result. A wrong code repeated for months distorts every decision built on it.

What do you do when self-reported data and analytics disagree?

Expect them to disagree, and treat the gap as information. Analytics might credit branded search while buyers credit a podcast. Both can be true: the podcast created awareness, and search was how they came back. Report both side by side rather than forcing one number to win.

The worst move is to pick one source and ignore the other. Teams that trust only analytics underinvest in channels that create demand. Teams that trust only self-reported answers can miss which pages and campaigns actually convert.

When two tools disagree on a number, the useful question is what each one is measuring. I wrote about handling that in what to do when two tools disagree about a number. The same thinking applies here.

What mistakes make this question useless?

The common mistakes are asking on every form, using a long dropdown, never reading the answers, and changing the categories too often. Each one either lowers answer quality or makes trends impossible to read. The fix is simple: one high intent form, open text, weekly coding, and a stable list of channels.

Another mistake is hiding the data. If answers live only inside the form tool, nobody looks at them. Pipe them into your CRM, whether that is HubSpot, Salesforce, or Attio, on the contact or deal record, so sales and marketing see them during normal work.

The last mistake is asking once and never connecting answers to revenue. The real value appears when you can say which first-touch sources produced closed deals, not just demo requests. That only works if the answer travels with the deal.

What should you do next?

Pick your highest intent form and replace any dropdown with an open text question plus a helpful hint. Add a coded channel field in your CRM, read and code answers every week, and review the results alongside your analytics each month. After one quarter, you will see sources your dashboards never showed.

This is a small change with a long payoff. It costs a few minutes a week, and it often shows founders that their best buyers come from places they were not paying attention to.

If you want help setting up the form, the CRM fields, and the automation that moves answers into your pipeline reports, reach out. I am happy to look at how you capture attribution today and suggest what I would change. Let's chat.